The social gauge

?If a friend asked you out for dinner at a restaurant they’d heard great things about, what would sway you – the fact that last night the restaurant served 900 people? Or the fact that three other friends with similar tastes had said how wonderful the food is?

In a study of 1,886 marketers conducted by MarketingSherpa in December 2008, 43% cited inability to measure ROI as a reason for rejecting a social media strategy. Sadly, it seems a determination to fit old measurement tactics to new forms of communication is leaving brands eating in the wrong restaurant. There should be no doubt that metrics and solid reports must be a pillar of any project. Simply put, if you don’t measure anything you don’t really know anything.

Online communities are no different: metrics are vital. Understanding the who, what, where, why and how many of your community is vital. Understanding if you’re doing your company some good (or bad) is vital. Setting KPIs is vital, and knowing whether you’re hitting them is vital. Metrics are vital.

To continue with the opening analogy, putting qualitative and quantitative measurements to the back of your mind (or worse, not considering them at all) is a little like setting up a restaurant, cooking a load of food, and not looking to see if anyone’s eating it. Recording, reporting and analysing your data is as much a part of managing an online community as keeping the spam out and the conversation going.

Where to begin
The real question is what to measure, why to measure it and what to do with the data you collect. There are some established standards when recording any web traffic, of course, all of which are numerical (numbers of hits, page views, pages per visit). Then there are some value-added, non-numerical data that form a standard of any web project’s measurement (most popular topics and pages, how people found your site, where people left your site).

Thanks to increasingly detailed web statistics software, it’s possible to drill down to very granular information. But that doesn’t necessarily mean that the data you find is of primary importance. Work out what you actually want to know, what is actually important to this project.

For a valid research community, getting a huge number of eyeballs on your content will not necessarily garner the insight you hope for. It can sometimes be more important to get detailed qualitative insight from ten people sharing their opinions than 10,000 people visiting the site, especially if they only read and don’t contribute anything.

Of course, aside from the traffic information there are also community-specific metrics, at the very least: number of members, number of active members, number of blogs/posts/comments/images. But it’s vital not to get caught in the zero-sum game – your community is not successful or unsuccessful because of bald numbers.

At FreshNetworks we’ve managed online research communities with many thousands of members, and some with just a hundred. Often the smaller communities better suit what clients want, delivering some of the most powerful insight available to brands, and far greater personal interaction than most websites can boast.

MarketingSherpa’s study concluded that social media lends itself to qualitative measurement, and that “marketers obsessed with only tracking social media results quantitatively are missing the point and may find themselves employing much less effective social media tactics for the sake of measurability”.

It also found that the difficulty of measuring the effectiveness of social media activity in traditional ways was putting companies off trying. Which is a real shame when you consider that social media platforms offer the chance to have genuine, meaningful conversations, rather than a series of anonymous click-throughs that may or may not foster any warmer feelings for the brand, or give the brands any insight.

Vital signs
Reporting on the current health and vitality of a community is more than just a numbers game, and ROI is more than just financial. Great stories from your community can form positive PR activity, and feedback (negative and positive) can inform improvements to customer services and spread learning about best practice throughout the company.

The numbers involved here are pretty much irrelevant, given that just one case study from a brand user could inform a whole PR campaign, or open up a whole new avenue of product development. The community manager – and let’s be clear on this, every community needs someone steering it – is the gatekeeper to all this knowledge. It is they who will record this feedback, filter it and report it. If it is collected in a meaningful way, it can be passed on to someone who can actually learn from it and investigate. It can be used to effect positive change.

It’s incredibly important to record negative sentiment within the community too. This is not rule-breaking, this is valuable insight. For many brands, this is the first time a customer or potential customer has come to them spelling out the problem and opening a dialogue. Far better to own this conversation, become aware of it immediately and have a chance to put it right, than to be blissfully unaware of any ill will, only to find yourself caught in an avalanche.

In 2005 Dell was the subject of an internet bushfire after blogger Jeff Jarvis complained about his “lemon of a computer” on his personal blog, BuzzMachine. Over the next few weeks, Jarvis’ blog was inundated with comments and talk of ‘Dell Hell’ was rampant across the blogosphere. At this stage Dell did not own the conversation and could not effectively measure the sheer volume of ill will.

At first the company stayed silent while its reputation burned, but eventually the debacle caused Dell to rethink its strategy. Dell’s people started to reach out to bloggers discussing their products, they tracked down complaints where they were happening, and offered to help. They set up IdeaStorm, a platform specifically for their customers to air their views, suggest ideas and improvements, allowing other customers to rate and vote on the strength of these ideas.

Somewhere in the Dell story, there will have been great swathes of bald numbers: not least the number of hits for ‘Dell sucks’ on Google, the number of comments on Jeff Jarvis’ blog, the number of complaints sent to the company, the dip in sales as a result. But the story, and the key for Dell, lay in what was said and how they could learn from it.

Thinking ahead
As with launching and moderating a community, monitoring stats and activity is not something to just have a go at and see what sticks. If you are serious about creating a valuable, worthwhile community, you need to think about recording and reporting metrics and activity before you’ve received even one visitor.

Planning is the key here. Any brand setting up in social media, whether that’s allowing a dedicated member of staff to control a Twitter account or launching a whole new community platform, must think clearly about why they’re doing it. Is it actually the right thing to do? What do you want to achieve and how will you recognise when you have achieved it? What will you do if sentiment is negative? What processes can you set up to make sure the right people know that they need to act?

Reporting should be based on these early thoughts. Because if you haven’t considered what you want from your community, no amount of measurement is going to tell you whether you have it.

Holly Seddon blogs at blog.freshnetworks.com. For more on metrics and reporting, see this post from June 2009.